The volume of data that the billions of things lying at the edge of the Internet of Things (IoT) will generate needs a comprehensive strategy for protocol mediation, processing, analysing, storing, securing, applying, and even sharing data to deliver value-creating and scalable use cases to industries and consumers. But, while trying to figure out how to manage “Big Data”, it is easy to forget that, when it comes to the IoT, it is actually the little things that matter. With the objective of maximising energy efficiency, many sensors are configured to only send small packets of data which, for convenience sake, can be called small data.
Small data is, for example, the temperature inside the storage area of a truck carrying perishable goods, sent out once every 10 minutes. It is the data relayed from a sensor placed above a parking bay that notes when the space is taken. Or the tiny packet of data relayed once a day from a water level sensor inside a reservoir, logging whether it is submerged or not. Small data gives context and enables us to identify patterns in behaviour, enabling machine learning, and driving data analytics, opening up a big world of opportunity.
The need for small data
The amount of data sent out in each case is minuscule – often no more than just a few bytes in size. And it needs to be, since larger data packets can place a heavy payload burden on the base station of a wireless IoT network that needs to connect and service millions of things. Hence, while it is easy to say “don’t sweat the small stuff”, for the IoT, it is the small stuff that truly matters. This is because all these small data packets eventually make up larger data sets from which to draw certain information. Take for example the parking-bay-sensor data mentioned earlier. Using the collected data from a parking lot over a period of months, the shopping centre management can trace a pattern of busy periods and quiet ones. This enables them to notify tenants when to run specials to attract more customers.
Even more important in future
As IoT usability expands, the reliance on small data packets that deliver more points of context become even more important. In certain use cases, a whole cascade of events will be triggered as soon as one sensor sends through specific data. One case in point is a patient being monitored at home – something we will see a lot more of as telehealth becomes more commonplace.
When accelerometer data from a wearable on an aged patient records an abrupt stop, it might indicate an injurious fall. This will trigger an automatic notification to the next of kin and the patient’s doctor. If no further movement from the patient is detected for a certain time, emergency services will be alerted to dispatch an ambulance. Furthermore, the patient’s smart home security system could also send out an access code once the ambulance crew arrives. This complete range of events is subject to the reliability of a small data sensor and a trustworthy network.
A network to depend on
To ensure the dependability of small data emanating from IoT sensors, SqwidNet, a subsidiary of DFA, is rolling out the SIGFOX IoT network in South Africa. The SqwidNet network is purpose built for listening to and delivering small packets of contextual data from these billions of connected things. Importantly, the SIGFOX standard ensures low-power usage, which is key to maximising sensors’ battery life.
Since our launch in November 2016, we have successfully deployed the network across all of South Africa’s eight major metros and we currently cover over 47% of the population. The network will exceed 85% of the population by the end of the year.
While the growth of IoT is a given, it is important to ensure that the foundations being laid now are stable and futureproof. To this end, securing small data’s place in the setup remains crucial. It its one small step for data; one giant leap for IoT.